Boost converter - input USB 5V, output about 6V

[Saimoun]

By ear I can clearly hear the difference between using the original design with the 3.3V LDO (which is dead silent) and using the charge pump (which has a bit of a white noise).

And just to be clear I do not have a PCB with the charge pump (yet), so for now the charge pump IC is simply soldered on a small adapter board and then attached to the original PCB (i.e. the one with the 3.3V LDO) using short cables.



EDIT: Nevermind, this noise is clearly unrelated to the charge pump as the exact same can be seen on both version (the quiet 3.3V LDO version and the charge pump version). I guess this could come from the USB power supply (external) which for sure has a switcher inside.
See next post for audio range noise floor measurements.

I tried to see on the oscilloscope but it's just general noise of about 50mVpp at high frequencies, the 1.2MHz switching freq is not at all predominant.

This was measured. It is very strange, the noise I can see is basically:
* The noise floor around 10mVpp
* A "clump" of 60mVpp noise at 10-11Mhz, of about 500ns, coming every 21us (=47kHz)

I've attached a few of the noise, the associated FFT, and a zoomed view of the "clumps".

[Saimoun]

So I tried to measure the noise on the audio range, since this is what matters to me. First I set the audio opamp in the circuit to add +30dB on the audio output.

There I can clearly see (and hear) the noise is louder, and that it is just white noise, there are no particular frequencies sticking out.


PS:  I am reaching the limitation of my cheap digital oscilloscope since the values are really low (about 2mVpp) so I am not sure how accurate this is.

[T3sl4co1l]

Could well be due to high frequencies being detected, resulting in variable offset levels at low frequencies.  But without a tight differential measurement I guess there's not much to say just yet?

Tim

[Saimoun]

Ok I understand - thanks for the help anyways Tim

Ironically it is a lot cheaper for me to order a new PCB and try the design properly rather than buying a proper oscilloscope so I'll do that